Newt Gingrich said in last Thursday's debate that if he became President, we'd have a base on the moon by the end of his second term. My immediate reaction was the same as Romney's: too expensive. But while Gingrich has some pretty wacky ideas, he's also a pretty smart guy, so I figured I ought to do the math.
The first sanity check was on the theoretical lower limit on the cost to send people to the moon. Barring a space elevator, that's going to involve rockets, which involves fuel. The Saturn V weighed 6,000,000 pounds, over 90% of which was fuel. Kerosene costs about $0.50 per pound, liquid hydrogen is more, and liquid oxygen less, so that might be a reasonable ballpark number for fuel. That tells us the theoretical lower limit is around $3,000,000 for three people, or $1,000,000 per person. That's expensive for the average person but it's not all that expensive in the greater scheme of things - there are commercial automobiles that cost more.
Of course, right now, the hardware costs a lot more than the fuel, and probably will continue to do so through the next decade, so that lower limit isn't reality. It does, however, justify a more detailed analysis.
Gingrich suggested devoting 10% of the NASA budget to prizes modeled on the Ansari X-prize for the first nongovernmental suborbital manned space flight. That prize resulted in a number of contenders, with a winner who is now working on a larger version for space tourism applications, for which several years' tickets have already been sold. The basic model seems reasonable, but do the numbers work out? Here are the numbers:
- The NASA budget for 2011 is $19 billion - around half a percent of the federal budget - so 10% of that is $1.9 billion.
- Virgin's Space Adventures sends space tourists into orbit, to the International Space Station, for $30 million each.
- Getting someone to the moon is roughly three times as expensive as getting them into orbit, both in terms of amount of fuel, and based on NASA's costs back in the 1960s.
That puts the per person cost of getting to the moon at around $90 million - and it also means that 10% of NASA's budget should be enough to send over 20 people to the moon each year, even without leverage from prizes or reductions in cost.
I draw two conclusions:
1. My gut reaction, and that of Mitt Romney, was wrong; Gingrich's moon base proposal isn't actually that expensive.
2. NASA is completely mismanaging the manned space program, given we haven't already done this. Perhaps this is why Gingrich wants to funnel the money to private players - they'll actually keep the costs down.
I think the cost to the moon is much more than 3x the cost to LEO. The Saturn V needed to launch 120,000 kg into LEO for the moon mission. The Soyuz capsule which has been used recently to get astronauts to the space station is only 5600 kg. Given that ratio you are looking at a 20x cost ratio.
I'm all for private enterprise working on cheaper spaceflight but I expect that Gingrich is getting a little carried away, at least in the short term.
The Wikipedia page for the Saturn V gives a payload to low earth orbit of 262,000 pounds, and to translunar injection of 100,000 pounds, which is where I got my factor of three. Granted that does not include descent to and ascent from the lunar surface, which seems to be another factor of two or so.
However, the original Soyuz at 7000 kg was a direct competitor to the Apollo command and service module combination, which was 30,000 kg, so a lot of that difference is just that Apollo was a gold plated inefficient design typical of NASA. The increase in Soyuz passenger capacity from 2 to 3 and perhaps any decrease in mass can probably be attributed to technological improvement since then.
So maybe it's 10 people to the moon per year with current technology instead of 20; that's still a lot.
Not that the details are particularly critical with the accuracy level we are shooting for here, but don't forget the extra radiation shielding, life-support, etc. There really is more equipment that you need to haul around to get to the moon versus just staying in LEO. Although I'm sure that NASA spends far more on comparable items than the Russians do, I don't think they are building them 4-5x heavier.
Of course the nice thing about a prize is that you can just set the award at 60 million per person and see whether anyone gives it a shot. If everyone thinks it's really 600 million per person then probably nothing will happen, in which case you haven't lost much.
The Soyuz capsules were designed as a part of a lunar program, so those capsules do indeed account for those factors. So yes, NASA was building them 3x heavier (4x heavier, but for 1.5x as many people).
By the way, the approach to space shielding is to put in as little as possible. Effective shielding from cosmic rays would have to have thicknesses measured in feet, so they instead take the approach of trying to leave the high energy gamma rays untouched, since those high energy gammas have lower cross sections for interaction with the human body than would the lower energy products that would result from their interaction with limited amounts of shielding.
Do you have any references for weight difference being due to NASA's inefficient design?
Looking at the relative weights of the habitable sections, the Apollo command module was 5560 kg and the Soyuz service plus reentry modules combined weight is 4880 kg. That does make the Apollo module heavier but not 3x heavier.
Most of the Apollo weight was in the service module which weighed almost 25,000 kg and contained the propulsion systems. It provided a delta v of 2800 m/s compared to the Soyuz orbital module which provides a delta v of about 300 m/s. That probably accounts for most of the weight difference.
Separate from the weight issue, I think you've got a very good point that the cost for many space activities could come down substantially if the proper incentives were in place. One of our friends was talking with the SpaceX folks at a conference recently about how they are looking at qualifying commercial chips for spaceflight instead of buying space rated chips. At least in cases where off-the-shelf technology can be repurposed for space it seems like you could easily get a huge cost improvement.
I think the Apollo service module is closer to the Soyuz orbital module, but it's probably not a perfect match either way since one spacecraft is divided into two components and the other is divided into three components.
So, let's say we take both of the Apollo modules together and all three of the Soyuz modules together. Do you have any references for the weight difference being due to NASA's inefficient design, as opposed to the 10x difference in delta v? I haven't found anyone claiming that in my quick Google searching but I'm open to being convinced.
My information was learned during the Apollo program in the 1960s and early 1970s. Basically everything was hugely overdesigned and thus inefficient.
While it doesn't directly address your point, here's an interesting article on the more efficient competing GE design that was rejected, and its similarity to the Soyuz:
But seriously, the Apollo team was willing to devote a 50 degree segment of the service module to, not just empty space, but ballast. Does that sound like an approach that cares about efficiency?
Yeah, I actually read that article last night since it was the first Google search result from the query, "Apollo vs Soyuz weight".
It does sound like there were plenty of elements which could have been done better. Some of the decisions were clearly political (choosing to use the third-best design according to NASA's own analysis). Other decisions such as having a section initially devoted to ballast (it was filled with equipment in one of the later missions) could be viewed as good forward thinking (I don't know the details of how that decision was made).
I guess the "gold plated inefficient design typical of NASA" accusation seems overly broad without a lot more detailed information. I would be more comfortable with "some aspects of the Apollo spacecraft could have been further optimized with respect to weight" to avoid bringing all of NASA into the discussion and to leave open the possibility that some of the weight non-optimality was due to emphasis on schedule or on features that the Soyuz was lacking (delta v, for example).
Actually, Obama has moved pretty far in this direction by historical standards, if only because NASA's own rocket program has collapsed into a singularity of Powerpoints. SpaceX looks to have a legitimate shot at taking over a huge chunk of the launch business, and their medium-term roadmap puts prices at about 10x that limit IIRC. (Although I understand the SpaceX Founder, Elon Musk, prefers a Mars base.)
I agree that the privatization of boost capacity looks set to happen anyway, which I regard as a good thing. Seen in that light, Gingrich's proposal is more a realization that the changes that are happening could allow a moon base to happen by 2020 without too much trouble.
On February 3rd, 2012 06:17 am (UTC), (Anonymous) commented:
(This is Tom N. replying.)
You hit the nail on the head with conclusion #2.
In terms of the costs to the moon, look at it this way: the costs of most major forms of transportation are a small multiple of their energy (fuel) costs. E.g., an airplane ticket probably costs you on the order of 3x the fuel cost.
Rockets currently cost something like 20,000x the fuel cost. So there is a _massive_ potential to reduce costs to orbit and then to elsewhere (e.g., the Moon).
As someone who is leading a start-up company that got its boost (capital and publicity) from winning a NASA technology prize, I can't say enough about how useful prizes can be, _IF_ they are done properly.
I'm confused why everyone here is using cost estimates of the recurring expense of sending someone to the moon to determine if it's too expensive to build a moon base.
I'm pretty sure what Romney is reacting to is the non-recurring cost of actually you know, inventing a moon base and a way to get there. The cost of the fuel is not even distinguishable in that number. You should be ratioing off of costs to design, build, launch, assemble, and maintain the ISS.
I agree that the commercial space industry is going to revolutionize the market. It's already in progress, but it's going to be awhile before the public notices.
According to whatever bogus numbers I can find in a 20 second google search, it costs $450M to launch the shuttle. The fuel cost is not more than $2M. My recollection is that Space-X is going to charge in the $40M's per launch. I think they will be paying much less than $2M per launch for their fuel (I have no reference here).
The cost of fuel is not a good measure of the cost for a space mission, even on a recurring basis, whether we are talking about commercial or government programs.
The cost of fuel was used, by me and by Tom Nugent, as a lower bound. My actual estimate is based on a transport cost a couple orders of magnitude more than fuel cost.
I think the costs of inventing and building a moon base efficiently would be much less than the costs of getting it there. You are correct that the cost of getting it there is currently a high multiple of the fuel cost.
Tom's point, and my assumption, is that the cost of transportation in space, while currently a huge multiple of fuel costs, ought instead to be a small multiple of fuel costs. That's also the direction everyone seems to agree things are going in.